Agentic Commerce: From Hype to a Multi-Trillion-Dollar Reality
Agentic commerce is rapidly emerging as the next structural shift in retail. A new whitepaper from Publicis Sapient and Salesforce, "Agentic Commerce: Unlocking the Emerging Shopping Experience", projects that AI agent-orchestrated commerce could generate between USD 3 trillion (approx. RM13.8 trillion) and USD 5 trillion (approx. RM23.0 trillion) in global revenue by 2030. Unlike previous digital waves, this transformation is not years away—it is already visible in how customers search and shop. The report notes that 44 percent of users who have tried AI-powered search now prefer it as their primary source, while platform agents such as ChatGPT already handle about 350 million shopping-related queries weekly. Traffic from generative AI browsers and chat services to retail sites has surged 4,700 percent year-over-year, underscoring that autonomous commerce is no longer theoretical. For retailers, this signals an urgent need to treat AI agents as first-class participants in the shopping ecosystem, not experimental add-ons.
Why Agentic Commerce Is Reshaping the AI Shopping Experience
Agentic commerce changes the basic mechanics of how demand is created, captured, and converted. Instead of customers manually searching, comparing, and purchasing, AI agents increasingly act on their behalf—interpreting intent, orchestrating options, and executing purchases autonomously. Publicis Sapient describes this as a structural shift: brands must now optimise experiences not only for human shoppers, but also for the retail AI agents that mediate those interactions. These agents are not confined to a single channel; they operate across the existing digital infrastructure—APIs, payment rails, logistics networks—so adoption is far faster than the internet or mobile rollouts. As a result, traditional funnel thinking becomes less relevant. What matters is how easily agents can discover products, understand policies, access live pricing and inventory, and complete secure transactions. Retailers that fail to become “agent-readable” risk disappearing from AI-curated consideration sets, even if their human-facing experiences remain strong.
The Competitive Edge for Early Movers in Autonomous Commerce
Salesforce’s Mohammed AlKhothani characterises agentic commerce as compressing a decade of transformation into two to three years. Early movers, he argues, will accumulate a durable edge in data, attribution, and integration that will be “exponentially harder to replicate” once the market consolidates. This advantage stems from becoming a trusted, high-signal source within AI agents’ knowledge graphs. As platform agents like ChatGPT, Google Gemini and others refine which retailers they recommend, integrated brands gain disproportionate visibility and traffic. At the same time, companies that wait risk being locked out by default preferences and established agent relationships. The launch of the Universal Commerce Protocol in January 2026—backed by major payments and retail players—further raises the stakes, setting an open standard for agentic transactions. Retailers that plug into this standard early are more likely to be recognised as reliable endpoints for autonomous purchasing and fulfilment, reinforcing their long-term competitive position.
Designing for Platform, Brand and Personal Retail AI Agents
The whitepaper highlights three key categories of retail AI agents that will shape the AI shopping experience. Platform agents, such as ChatGPT and Google Gemini, sit above individual brands and aggregate options across the web, often becoming the first touchpoint in a customer journey. Brand-owned agents operate within enterprise ecosystems—websites, apps, and service channels—using proprietary data to drive personalised recommendations, support and transactions. Personal consumer agents act directly on behalf of individuals, executing tasks like replenishment or complex purchase comparisons with minimal human input. Retailers must understand how their product data, content, and policies surface differently in each context. That means exposing machine-readable catalogues, clear service-level commitments, and rich metadata that agents can parse. It also demands robust governance around security, consent, and compliance as agents exchange sensitive identity, payment and preference data across multiple platforms and ecosystems.
Building an Agent-Ready Retail Stack: The A.C.E. Framework
To help retailers operationalise autonomous commerce, the report introduces the A.C.E. Framework—a three-layer model for building agent-ready ecosystems. The Agentic Experience Interface focuses on making products and services discoverable and actionable by AI agents, emphasising standards-based APIs and structured content. Composable Micro-Apps expose modular commerce capabilities, such as pricing, inventory, and checkout, as interoperable services that agents can invoke on demand. Enterprise Context Orchestration sits beneath, governing data, security, compliance, and contextual intelligence across all interactions. Crucially, the whitepaper warns against treating generative AI as a simple plug-in: MIT NANDA research cited in the report finds that 95 percent of initiatives fail to generate measurable impact, often due to workflow misfit, weak memory architecture, and missing service-level objectives. A phased implementation roadmap—moving from pilot to production in around 90 days—aims to help retailers avoid these pitfalls and turn experimentation into scalable value.